I've experience working in academia and the private sector... I don't think it's a matter of competition, universities simply aren't pushing students: everything is spoon-fed, there are very few lecturers who would say "go learn about X".
With an academic hat on I can see the advantage of staying with theoretical topics - teach the basis well and it is applicable to any language or environment. But universities are struggling to stay relevant (and afloat in our budget-constrained times). With corporate research outstripping university research because of the decreasing academic appetite for risk, universities need to be moving with the times, not retreating into maximising student throughput and grant money - teaching essential job skills for programming doesn't have to be mutually exclusive with computer science theory.
There's this bizarre focus on single languages - previously Java and now C#... and spending a lot of time teaching them to students. That runs the risk of the student only learning skin-deep how principles are applied in Java or C#. It's not really fair to compare MIT to another random university but watching their Open Courseware videos it's clear how much those students are expected to figure out on their own - all universities should expect that from their students (because if you're not good at programming you shouldn't be in a programming degree and you *definitely* shouldn't be passing it).
Students should be pushed to learn languages on their own, not spending an entire course learning a language - by the time they graduate they'd better be familiar with a load of languages which force them to think differently about their solutions.
My primary concern is that there's very little focus on letting students learn wisdom about refactoring/good design/etc because they never live with their code - they don't have to deal with the crappy code they wrote 6 months ago so they don't learn the benefits of doing it right the first time (or of realising you made a mistake and rewriting and refactoring it).
Just dipping in here but I think the poster's point was not that a Dollar has inherent value, but that a large government with a lot of power is prepared to enforce the use of the Dollar for the payment of all debts within the borders of its country.
Personally (and this isn't very extensively thought out because I don't really care about this topic in the grand scheme of things) I don't know how useful something like Gold is as a currency - yes, it has some inherent value for manufacturing but if the global economy crashes (which seems to be the argument of people who want to use gold for everything) then surely crops and livestock are what people will want to trade for? And Gold isn't that useful in creating ploughs...
oops, that is, they describe the password reset feature on their website http://helpdesk.lastpass.com/account-recovery/ and it's not a simple "confirm your identity and we'll e-mail you a new password" system
While that would be nice to know I don't think it's relevant to a postmortem: they described the architectural elements which encountered the failures.
FYI, though, based on what they've said today and in the past: it seems that they are using regular servers with disks rather than a SAN & I believe they use GNBD to connect the EBS server disk images and EC2 instances (rather than iSCSI)
Yeah, I think that's what they're trying to do. I suppose it makes sense in a way, they want to make sure load is evenly distributed across their availability zones . But it seems to me they could have prevented that through better API design (e.g. users expressing a constraint that 2 resources should be in the same zone where that's meaningful but otherwise not permitting the selection of a specific zone)
I believe they generate a new HTML document each time a comment is added or up/downvoted - they could replicate the comment and vote data to another site.
It'd be an increase in traffic but not necessarily a huge increase in load (since they wouldn't be generating HTML at the second site unless they're in failover mode).
I don't know whether the increased reliability would be worth the extra load in their case, however, since I doubt they lose that much money from downtime (given how frequently they're down)
Amazon have complete isolation between Regions and good isolation between Availability Zones.
At work we'd recommend people use 2 cloud providers for their important services (which could be 2 Amazon regions or it could be Amazon and Rackspace) to prevent this sort of failure taking your business offline. You can't rely on any particular cloud provider to be reliable but it's a reasonably safe bet that a selection of cloud providers won't have significant overlapping downtime
It's also worth pointing out that all cloud SLAs are basically useless: if Amazon falls below their advertised uptime they'll refund you some of your charges - but they'll never refund more than what you've paid them: they don't compensate you for all the money you're losing (and the AWS charges are likely pocket change compared to this)
Their API gives different names for the availability zones for each user (so your us-east-1d could be my us-east-1a) which complicates talking about issues (since all you can say is "two availability zones are experiencing problems"), especially when your system uses multiple accounts
Jokes aside, if people use The Cloud (I'm using this tongue in cheek...) rather than a cloud this thing doesn't happen.
We use a number of providers which means that even if Amazon fell over completely our systems would be fine -- it looks like a lot of sites (reddit, for instance) don't bother to do this.
Amazon's Availability Zones are designed to have separate power, cooling and network so I don't think this is the issue.
It was (is) a problem with their disk subsystem in multiple availability zones so I suspect they were in the process of pushing out some new storage controller code and some bug didn't appear until the later stages of their rollout.
From their status log it looks like they're manually correcting the issue with each disk.
I agree completely, I was just making the point that it is a bit more complicated than X forwarding. As I said, I don't think this should be patentable -- specifically, I think it doesn't go into enough detail on the methods used to ameliorate latency and bandwidth issues.
I don't think this should be patentable but in my mind the implementation of OnLive is doing something a lot more complicated than X forwarding (the idea, of course, is the same).
It's working over a contended public network with a large number of hops with varying latencies, doing low latency compression on HD resolutions at reasonably low bandwidth.
As I understand it, wifi adds more unreliability (packet loss and latency) to the network path as far as they're concerned
I use LastPass which lets me store all my passwords and has an iPhone app for when I'm out and about (and if I'm more trusting, a client-side-decryption web-based system).
The only annoyance I have is when I need to set up an IMAP account (and then I just bring up the password for that account on my phone and type it in). You get all of the functionality (except the iPhone app) free.
It's a much better compromise for me... I keep non-web passwords too (like router passwords for when I need to go down to the datacentre)
While I'll wait until it's gone through peer review, I'm sure this means that news outlets will pick this up and misreport it... it's good PR for HP's research group either way!
I think you'll find he was referencing the contentious area of tabs vs. spaces (and if spaces, how many) as being a perfect way to keep a roomful of programmers busy for hours (reinforced by the second example, "should I avoid braces if I only have a single statement?".
I was using the accidental use of "newline" by the poster as an excuse for a small joke.
They've open-sourced management code from their Infrastructure-as-a-Service cloud provider (like Amazon EC2) and their blob store (like Amazon S3), meaning you can use this release to build a public/private cloud provider.
The IaaS provider code ("OpenStack Compute") is a preview and they say October is their release date.
As an open source cloud provider codebase it joins Eucalyptus, OpenNebula and a few others -- however (and since I've not used the rackspace cloud I don't know whether this is at all accurate) it should be much more stable since it's most of the code running a big cloud provider (whereas Eucalyptus and OpenNebula still tend to be rather flakey in my experience).
What's not entirely clear is how much they're holding back from the release - they say OpenStack Compute is "the same technology that underlies two of the largest and best ones out there". If they push features over to the open source project to keep parity with their own implementation then it may gain (or, in fact, may already have - I've not looked into the code much) features such as energy efficient host cluster management, shutting down VM hosts when they're not required.
There isn't really a major risk at the minute to doing this - the technology behind cloud providers is pretty simple, the cost of running a public cloud is buying and continually refreshing all the hardware and getting plenty of bandwidth and datacentre space to run workloads. Anyone with the money to spend on the hardware either probably doesn't mind spending some developer time building their own API (especially since if their customers use the API directly then having your own API is a good way to discourage customers from moving to another provider)
To maneuver you simply make components of the sail non-reflective, as I understand it... it won't give you a fast turn but as long as we don't want to dogfight in space it should be acceptable
We get a much better bang for our buck by using commodity hardware - we can build & maintain a few linux SANs for the price of an 'enterprise' one. We just build software which can deal with the underlying hardware failing or the network splitting.
Slashdot Mirror
Your signature is amazingly appropriate for the content of your message :-)
What if I said "i need to find a cheap glockenspiel"
I just tried this - it performed the right search for me - neat!
You tell siri about your relationships: "Siri, Alice is my aunt", "Siri, Bob is my boyfriend", etc.
I've experience working in academia and the private sector... I don't think it's a matter of competition, universities simply aren't pushing students: everything is spoon-fed, there are very few lecturers who would say "go learn about X".
With an academic hat on I can see the advantage of staying with theoretical topics - teach the basis well and it is applicable to any language or environment. But universities are struggling to stay relevant (and afloat in our budget-constrained times). With corporate research outstripping university research because of the decreasing academic appetite for risk, universities need to be moving with the times, not retreating into maximising student throughput and grant money - teaching essential job skills for programming doesn't have to be mutually exclusive with computer science theory.
There's this bizarre focus on single languages - previously Java and now C#... and spending a lot of time teaching them to students. That runs the risk of the student only learning skin-deep how principles are applied in Java or C#. It's not really fair to compare MIT to another random university but watching their Open Courseware videos it's clear how much those students are expected to figure out on their own - all universities should expect that from their students (because if you're not good at programming you shouldn't be in a programming degree and you *definitely* shouldn't be passing it).
Students should be pushed to learn languages on their own, not spending an entire course learning a language - by the time they graduate they'd better be familiar with a load of languages which force them to think differently about their solutions.
My primary concern is that there's very little focus on letting students learn wisdom about refactoring/good design/etc because they never live with their code - they don't have to deal with the crappy code they wrote 6 months ago so they don't learn the benefits of doing it right the first time (or of realising you made a mistake and rewriting and refactoring it).
Just dipping in here but I think the poster's point was not that a Dollar has inherent value, but that a large government with a lot of power is prepared to enforce the use of the Dollar for the payment of all debts within the borders of its country. Personally (and this isn't very extensively thought out because I don't really care about this topic in the grand scheme of things) I don't know how useful something like Gold is as a currency - yes, it has some inherent value for manufacturing but if the global economy crashes (which seems to be the argument of people who want to use gold for everything) then surely crops and livestock are what people will want to trade for? And Gold isn't that useful in creating ploughs...
oops, that is, they describe the password reset feature on their website http://helpdesk.lastpass.com/account-recovery/ and it's not a simple "confirm your identity and we'll e-mail you a new password" system
They describe the password reset feature. Another post also gave a good interpretation of how it likely works
LastPass said that the level of traffic they saw in the attack was enough for the password hashes + salts but not enough for users encrypted blobs.
While that would be nice to know I don't think it's relevant to a postmortem: they described the architectural elements which encountered the failures.
FYI, though, based on what they've said today and in the past: it seems that they are using regular servers with disks rather than a SAN & I believe they use GNBD to connect the EBS server disk images and EC2 instances (rather than iSCSI)
Yeah, I think that's what they're trying to do. I suppose it makes sense in a way, they want to make sure load is evenly distributed across their availability zones . But it seems to me they could have prevented that through better API design (e.g. users expressing a constraint that 2 resources should be in the same zone where that's meaningful but otherwise not permitting the selection of a specific zone)
I believe they generate a new HTML document each time a comment is added or up/downvoted - they could replicate the comment and vote data to another site.
It'd be an increase in traffic but not necessarily a huge increase in load (since they wouldn't be generating HTML at the second site unless they're in failover mode).
I don't know whether the increased reliability would be worth the extra load in their case, however, since I doubt they lose that much money from downtime (given how frequently they're down)
Amazon have complete isolation between Regions and good isolation between Availability Zones.
At work we'd recommend people use 2 cloud providers for their important services (which could be 2 Amazon regions or it could be Amazon and Rackspace) to prevent this sort of failure taking your business offline. You can't rely on any particular cloud provider to be reliable but it's a reasonably safe bet that a selection of cloud providers won't have significant overlapping downtime
It's also worth pointing out that all cloud SLAs are basically useless: if Amazon falls below their advertised uptime they'll refund you some of your charges - but they'll never refund more than what you've paid them: they don't compensate you for all the money you're losing (and the AWS charges are likely pocket change compared to this)
Their API gives different names for the availability zones for each user (so your us-east-1d could be my us-east-1a) which complicates talking about issues (since all you can say is "two availability zones are experiencing problems"), especially when your system uses multiple accounts
Jokes aside, if people use The Cloud (I'm using this tongue in cheek...) rather than a cloud this thing doesn't happen.
We use a number of providers which means that even if Amazon fell over completely our systems would be fine -- it looks like a lot of sites (reddit, for instance) don't bother to do this.
Amazon's Availability Zones are designed to have separate power, cooling and network so I don't think this is the issue. It was (is) a problem with their disk subsystem in multiple availability zones so I suspect they were in the process of pushing out some new storage controller code and some bug didn't appear until the later stages of their rollout. From their status log it looks like they're manually correcting the issue with each disk.
I agree completely, I was just making the point that it is a bit more complicated than X forwarding. As I said, I don't think this should be patentable -- specifically, I think it doesn't go into enough detail on the methods used to ameliorate latency and bandwidth issues.
I don't think this should be patentable but in my mind the implementation of OnLive is doing something a lot more complicated than X forwarding (the idea, of course, is the same).
It's working over a contended public network with a large number of hops with varying latencies, doing low latency compression on HD resolutions at reasonably low bandwidth.
As I understand it, wifi adds more unreliability (packet loss and latency) to the network path as far as they're concerned
I use LastPass which lets me store all my passwords and has an iPhone app for when I'm out and about (and if I'm more trusting, a client-side-decryption web-based system).
The only annoyance I have is when I need to set up an IMAP account (and then I just bring up the password for that account on my phone and type it in). You get all of the functionality (except the iPhone app) free.
It's a much better compromise for me... I keep non-web passwords too (like router passwords for when I need to go down to the datacentre)
While I'll wait until it's gone through peer review, I'm sure this means that news outlets will pick this up and misreport it... it's good PR for HP's research group either way!
I think you'll find he was referencing the contentious area of tabs vs. spaces (and if spaces, how many) as being a perfect way to keep a roomful of programmers busy for hours (reinforced by the second example, "should I avoid braces if I only have a single statement?".
I was using the accidental use of "newline" by the poster as an excuse for a small joke.
Ask "Should I use spaces or tabs for newlines?"
Call me traditional but I like to use newlines for newlines :-P/p>
They've open-sourced management code from their Infrastructure-as-a-Service cloud provider (like Amazon EC2) and their blob store (like Amazon S3), meaning you can use this release to build a public/private cloud provider. The IaaS provider code ("OpenStack Compute") is a preview and they say October is their release date.
As an open source cloud provider codebase it joins Eucalyptus, OpenNebula and a few others -- however (and since I've not used the rackspace cloud I don't know whether this is at all accurate) it should be much more stable since it's most of the code running a big cloud provider (whereas Eucalyptus and OpenNebula still tend to be rather flakey in my experience).
What's not entirely clear is how much they're holding back from the release - they say OpenStack Compute is "the same technology that underlies two of the largest and best ones out there". If they push features over to the open source project to keep parity with their own implementation then it may gain (or, in fact, may already have - I've not looked into the code much) features such as energy efficient host cluster management, shutting down VM hosts when they're not required.
There isn't really a major risk at the minute to doing this - the technology behind cloud providers is pretty simple, the cost of running a public cloud is buying and continually refreshing all the hardware and getting plenty of bandwidth and datacentre space to run workloads. Anyone with the money to spend on the hardware either probably doesn't mind spending some developer time building their own API (especially since if their customers use the API directly then having your own API is a good way to discourage customers from moving to another provider)
To maneuver you simply make components of the sail non-reflective, as I understand it... it won't give you a fast turn but as long as we don't want to dogfight in space it should be acceptable
(Hungary, I think?)
We get a much better bang for our buck by using commodity hardware - we can build & maintain a few linux SANs for the price of an 'enterprise' one. We just build software which can deal with the underlying hardware failing or the network splitting.